1 / 36

A CP in Charm Decays at LHCb

A CP in Charm Decays at LHCb. B. Viaud (LAL-in2p3) On behalf of the LHCb collaboration. Why Charm Physics ?. Charm Physics is essentially a 2-generation physics: any CPV above O(0.1%) means something new.  NP, or unexpected strong effects.

guinevere-burch
Download Presentation

A CP in Charm Decays at LHCb

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. ACP in Charm Decays at LHCb B. Viaud (LAL-in2p3) On behalf of the LHCb collaboration

  2. Why Charm Physics ? • Charm Physics is essentially a 2-generation physics: any CPV above O(0.1%) means something new.  NP, or unexpected strong effects • D-D mixing, CP violating decays and rare decays involve FCNC’s that are strongly GIM-suppressed (low mass down-type quarks in the loop)  NP contributions can have measurable effects (not hidden by SM) • FCNC with down-type quarks in the loop: constrain NP couplings that • can’t be reached by B/K decays.  Complementarity with the B-physics program. • Very large samples of charmed particles at hadronic colliders !  Charm decays are a good place to look for NP and constrain its properties !

  3. CP Violation in Charm Two complementary ways to seek CPV (and NP) in Charm Decays • D oscillate, so one can look for two manifestations of indirect CPV - CPV in mixing: D0  D0  D0D0 - CPV in the interplay between mixing and decay • A(Df)A(Df):direct CPV • Direct CPV is as good an opportunity as indirect - Mixing is slow, strong phases can be large in decays. - While indirect CPV is nearly universal, direct depends a lot on the final state. Measuring many brings many complementary clues. • CPV is small: ~0.1% to ~1% for direct CPV  What’s SM; What’s NP ? Probably an order of magnitude below for indirect CPV. Today: direct CPV @ LHCb. Focus on the current most precise example: Acp(KK)-Acp()

  4. LHCb

  5. NP via the precision study of CPV and Flavor Physics LHCb • Key point: huge b and c production in high E p-p collisions - @s=7 TeV: (ppbb+X)=(284 ± 20 ± 49) μb [1] (ppcc+X)=(6100 ± 930) μb [2] In 1fb-1: ~1012 cc pairs in LHCb’s acceptance • Key point: dedicated experiment, optimized for Flavor Physics • in a hadronic environment. - Forward detector - Performant vertexing, p and M reconstruction, particle-ID - Very selective, polyvalent and configurable trigger. [1] Phys. Lett. B694: 209-216, 2010 [2] LHCb-CONF-2010-013

  6. B-field polarity can be reversed: Up or Down • Typical Performance • Charged tracks momentum: p/p=0.35-0.55%, m=10-20 MeV/c2 • ECAL: E/E=10%/E  1% (E in GeV) • muon-ID () ~95%, mis-ID rate()~1% • K- separation(KK) ~95%, mis-ID rate(K)~10% • Proper time: t~ 30-50 fs, z~ 60m (Prim. Vtx) z~ 150 m (Secondary Vtx)

  7. Trigger/DAQ • Trigger Efficiency • ~30% for multibody hadronic modes • ~90% for di-muons Ex 1/fb : High PT candidates ~0.5M D0+-+- on tape • Output Rate High PT displaced tracks matched with L0 objects. • 3-4 kHz in 2011 • 4.5 kHz in 2012 Full reconstruction (ex: lifetimes) Inclusive/exclusive modes Highly configurable. Easy to add/remove/prescale modes. 3-4 kHz

  8. Peak Luminosity • 2011: 3-4 1032/cm2/s • 2012: 4 1032/cm2/s • <#collisions> per bunch crossing ~1.5 “Luminosity Leveling” to obtained that from LHC’s luminosity 2011 2012 ~1.0 fb-1 @ s=7 TeV 0.3 fb-1 @ s=8 TeV

  9. Peak Luminosity • 2011: 3-4 1032/cm2/s • 2012: 4 1032/cm2/s • <#collisions> per bunch crossing ~1.5 “Luminosity Leveling” to obtained that from LHC’s luminosity 2011 2012 Hope to record 1.5 fb-1 in 2012 + 2.5 fb-1 in 2015/2016 ------------------------------ Ltot=5fb-1

  10. ACP= ACP(D0K+K-)-ACP(D0+-) - 0.6 fb-1 (2011) - Phys.Rev.Lett. 108 (2012) 111602

  11. Analysis Strategy Use D*: Qslow  tells D0’s flavor • Measure a time-integrated asymmetry • First order Taylor Expansion: When f=+- or K+K-: no detection asymmetry between D and D  AD(f)=0 Similar for f=+- and K+K- ARAW = ARAW(K+K-) - ARAW(+- ) = ACP

  12. ARAW = ARAW(K+K-) - ARAW(+- ) = ACP • This rule gives a very robust way to detect a CPV effect • But remember ! It can be broken by • Large asymmetries (>>1%): Taylor Expansion breaking down • Dependence of AP(D*) and AD(s) upon (KK)/(). • Ex: AD(s) depends upon the Sphase space, and • KK and  selections favor a different region. • Different and asymmetric peaking backgrounds. • So the fun in this analysis is to avoid those problems. • Main protections: • Measurements in separate bins of PT and  of D*’s, P of S • Fiducial cuts to remove regions of large asymmetry • Many checks…

  13. What does ACP measure exactly ? • Time integrated asymmetries: a combination of direct & indirect CPV. Depends on <t> of the D0 in the sample (~time given the mixing to interfere). • Indirect CPV universal to a very good approximation, but lifetime acceptance can differ between KK and .  Also measure <t> to disentangle each contribution • A year ago…

  14. Selection Cut-based selection: use the decay topology and kinematics, and LHCb’s PID performance. • Track & Vertex fit quality • Tracks must not come from the • primary vertex (PV) & ct(D)>100 m. • D must come from the PV, to reject D* from B decays •  between D0 in lab frame and its daughters • in D0 rest frame: |cos|<0.9 • Tracks identified as kaon/pions using PID info from the RICH • PT(D)>2 GeV/c N.B. This offline selection applied on candidates that fired a similar (looser) selection in the High Level trigger

  15. Fiducial cuts The magnetic field breaks the symmetry of the detector Horizontal plane  Kinematic regions where ARAW can reach 100% ! Borders where +/- are swept out while -/+ are swepted in. (this includes also the beam pipe)

  16. Fiducial cuts ARAW for |py/pz| slow < 0.2. Horizontal plane. Kinematic regions where ARAW can reach 100% ! • Breaks the formalism (too large an • for a Taylor expansion) • Possible second order effects if the efficiency for being in this region differs between KK and . • Depends more on PX than on PT,D*, D* • or Pslow  Directly in beam pipe Thus: not treated perfectly by the kine. binning Py • Left-right binning + the fact that ~1/2 the sample is taken with B-field Up and ~1/2 with • B-field Down should limit the overall effect. • However, to be more robust, sacrifice 25% of • the statistics with Fiducial cuts Px

  17. Fiducial cuts ARAW for |py/pz| slow < 0.2. Horizontal plane. Kinematic regions where ARAW can reach 100% ! • Breaks the formalism (too large an • for a Taylor expansion) • Possible second order effects if the efficiency for being in this region differs between KK and . • Depends more on PX than on PT,D*, D* • or Pslow  Cross the beam pipe after VELO / before T-stations Thus: not treated perfectly by the kine. binning Py • Left-right binning + the fact that ~1/2 the sample is taken with B-field Up and ~1/2 with • B-field Down should limit the overall effect. • However, to be more robust, sacrifice 25% of • the statistics with Fiducial cuts Px

  18. Mass spectra and signal yields

  19. Signal Extraction 1 Bin • In 216 bins 54 bins in PT,D*  D*  Pslow  left/right  2 Mag Up / Mag Down  2 Before/After an LHC technical stop 1 2 • Fit to m distributions Signal: double gaussian convolved with a function describing a asymmetric tail. 1 D*+ and D*- parameters float separately. Background: B[ 1 - exp( -(m- m0)/C ) ] 2 • Finally: ARAW and ARAW in each bin, then weighted average ACP =(-0.82  0.21stat )% (2 / NDF = 211/215) • Fit to background subtracted decay time distributions yields:  This would essentially be a direct CPV

  20. Systematics ACP =(-0.82  0.21stat  0.11)% 3.5  from no CPV.

  21. Cross Checks • Electron and muon vetoes on the soft pion and D0 daughters • Different kinematic binnings • Stability of result vs data-taking runs • Stability vs kinematic variables • Toy MC studies of fit procedure, statistical errors • Tightening of PID cuts on D0 daughters • Tightening of kinematic cuts • Variation with event track multiplicity • Use of other signal, background line-shapes in the fit • Use of alternative offline processing (skimming/stripping) • Internal consistency between subsamples (splitting left/right, field up/ field down)

  22. Cross Checks • No evidence of dependence on relevant kinematic variables

  23. Stability wrt PID • Stability with time No significant variation of ACP when tightening the cut on the hadron PID information provided by the RICH PID tight+ Final value ACP =(-0.88  0.26stat )% PID tight++ ACP =(-1.03  0.31stat )% A technical stop occurred here • Internal consistency: • a closer look Split the 216 bins into 8 smaller sets and check 2 for each, and betweenthem: 2 / NDF = 6.7/7

  24. World Wide CDF public note 10784 Agreement with no CPV: 610-5

  25. SM or NP ?? • Predictions are difficult with D mesons - Too light (heavy) for the techniques that work in B (K) physics • Present consensus - Difficult for the SM to generate more than O(10-4-10-3) (canonic point of view till 2011) - But possible: one can think of Hadronic enhancements pushing it up to O(1%) - Would help: Individual asymmetries - Would help: Several decay modes should be affected by the same NP, but not the same strong effects: compare ACP measured in each mode to distinguish enhanced contributions of higher order standard model diagrams from NP effects Ex:  D+(S) KSh+; h+  D+ K+K*0; K*+K0  D+ 0+; + 0 ; +’  DS K+, K+’ , K(*)0+  Dh+h-h+ ; h+h-h+h- Grossman, Kagan, Zupan (arXiv:1204.3557)

  26. Prospects Short term (1.1 or 2.5 fb-1) • Update ACP= ACP(K+K-)-ARAW(+-)  from 0.25% to ~0.15% may be enough to confirm a 4-5 effect. • ACP with D+(S) KSh+ vs. h+(work started !)  Expect ~7M D++ and ~3.5M D+KS+ Belle: ACP (D++ vs. D+(S)+) = (0.510.280.05)% with 0.238M D++ Belle: ACP (D+KS+) = (0.360.090.07)% with 1.7M events arXiv:1203.6409 CPV due to the kaon • Dalitz analyses of Dh+h-h-, h+h+h-h- modes Longer term: LHCb upgrade (2019)

  27. Conclusion ACP (K+K--+-)=(-0.82  0.21stat  0.11)% • Control of systematic effects: good ex. of precision physics @ pp collider. • Evidence for CPV in charm decays at LHCb  Mostly a direct CPV  Not yet a 5 effect  But not far from it when combined with other experiments (4) • Could be SM, could be NP, it’s anyway very interesting. • There’s a large Charm physics programme at LHCb. Other modes will be studied in the future to over-constrain the problem. • And don’t forget the LHCb’s upgrade !  Stay tuned (at least for the next 15 years  ) !

  28. Back-up

  29. LHC’s Schedule M.Nessi, Chamonix 2012

  30. Upgraded LHCb (start by 2019) Should bring ~180 times more hadronic charm decays ! • 50 fb-1 with Lpeak=1-21033 cm-2s-1 • At s=14 TeV: CC ~1.8 times larger • Fully software trigger: Trigger Efficiency on hadronic decays 2 (reduce the role the hardware L0 trigger) -This means ~460M D0K+K- & 130M D0+-. Naïve extrapolation: Acp~0.015%. That’s far below the current systematics. A part of the statistic could be sacrificed to improve it. (2011: 1MHz) -Also for decays like D+(S) KSh+ vs. h+, will we probably be pushing on the systematics by then. -And many other things: DCS, precision Dalitz studies, etc… See e.g. “Workshop on the Implications of of LHCb measurements”, CERN, April 16-18, 2012 (2011: 4kHz)

  31. Preliminary estimates ! + Not everything is solved by increasing the statistics. In some cases, some will be sacrificed to improve systematics.

  32. Fraction of indirect CP • Reminder: • <t>0 since the lifetime acceptance • differs between KK and  e.g. Smaller KK opening angle: easier to miss cut vetoing tracks from Primary Vertex. • Fit to background subtracted decay time distributions yields: Essentially due to the fraction of D* from B decays

  33. Fraction of indirect CP • Reminder: • <t>0 since the lifetime acceptance • differs between KK and  e.g. Smaller KK opening angle: easier to miss cut vetoing tracks from Primary Vertex. • Fit to background subtracted decay time distributions yields:  Indirect CPV mostly cancels

  34. Preliminary estimates !

More Related